-
Notifications
You must be signed in to change notification settings - Fork 394
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
guide: Sharing Experiments #2618
Conversation
This draft is almost ready in scope. Could you check the structure as a UG document? @jorgeorpinel @shcheklein I'll write a similar "Cleaning Up Experiments" before the new GS:Experiments #2574 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It looks good to me!
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
Okdk I did a pretty major copy edit of your work @iesahin but didn't touch the story, logic, or instructions. I think this is mergeable so feel free, but I do have some comments that we can address now or as another PR later ⏳ |
content/docs/user-guide/experiment-management/sharing-experiments.md
Outdated
Show resolved
Hide resolved
DVC has a mechanism called the [run-cache] that stores the artifacts from | ||
intermediate stages. For example, if there is an intermediate step that applies | ||
data-augmentation on your dataset and you would like to push these artifacts as | ||
well as the end products of the experiments, you can use `--run-cache` flag to | ||
push all of these to the DVC remote. `--run-cache` flag pushes all artifacts | ||
referenced in `dvc.lock` file. | ||
|
||
[run-cache]: /doc/user-guide/project-structure/internal-files#run-cache |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't think that's what the run-cache does here. Intermediate artifacts of the experiment should be moved by exp push/pull
. The run-cache probably only exists here as an edge case when you use run/repro
in between exp run
s. If so, no nee to even mention it probably.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
p.s. This is the only sort of major thing I would prefer to address before merge.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Oh and if we need to keep this part about the run-cache, what about pull --run-cache
? It's not mentioned in that section.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Frankly I didn't test --run-cache
option and this is what I got from the Ref. Need to test this I believe.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Removed for now. Well, I left a small note.
content/docs/user-guide/experiment-management/sharing-experiments.md
Outdated
Show resolved
Hide resolved
content/docs/user-guide/experiment-management/sharing-experiments.md
Outdated
Show resolved
Hide resolved
make example about exp push/pull --all per #2618 (review) and #2618 (review)
example about cloning individual experiments per #2618 (review)
@iesahin thanks for the updates. I did another round too. I think this can be merged (feel free) but there are some pending discussions ☝️ we can address separately if you prefer. Up to you — if there's no rush to merge let's keep iterating. |
This PR adds a new section/page to UG:Experiment Management.
Topics to be covered
Closes #2367
Moved todo to #2687